103M app events study: AI use narrows work fragmentation

A new study by Sumer S. Vaid and Ashley V. Whillans, analyzing 103 million application events from 1,017 knowledge workers across eight organizations (law, financial services, etc.), reveals that digital fragmentation—the constant switching between apps—is primarily driven by day-to-day fluctuations within individual employees. These internal variations accounted for 44.6% of fragmentation, versus 35.8% from stable personal differences and only 19.6% from organizational context. Fragmentation increased over the workweek and reset after weekends and holidays. Notably, higher-than-usual use of communication apps (email, Slack, Teams) correlated with more fragmented work patterns.

Generative AI use emerged as a double-edged sword: AI tools were more likely to be used on already fragmented days, but the period immediately following AI interaction showed a clear shift toward narrower, longer, and more predictable application use. This suggests that instead of merely compounding distraction, AI may help structure work by focusing attention after an AI-assisted task. The findings have direct implications for workplace design—suggesting that interventions aimed at reducing fragmentation should target daily routines rather than just individual habits or organizational policies, and that thoughtful integration of generative AI could serve as a cognitive anchor in chaotic workdays.